Users commonly provide video content to websites (e.g., YouTube®), which can be referred to as “posting a video.” The user can spend a significant amount of time to convey the message of the video before the user selects the video (e.g., by clicking the video displayed on a website). For example, the user can associate a title, a static thumbnail image, and/or a textual description with the video. Users often have a difficult time when the video originates on a different website and the user tries to upload their video to a video server. Further, the title may not be descriptive of the contents of the video, the static thumbnail image may not summarize the essence of the video, or the description of the video may be a poor signal for whether the video will be interesting to a viewer.
Video browsing is also limited. Other users (e.g., viewers) can access and view the video content via the websites. For example, the viewers can see a video's title and static thumbnail of the video before deciding whether to play the full video. However, the viewers may find it difficult to select particular videos of interest because the title may not be descriptive of the contents of the video, the static thumbnail image may not summarize the essence of the video, or the textual description with the video may be a poor signal for whether the video will be interesting to the viewer. Thus, the viewers may spend significant amounts of time searching and watching videos that are not enjoyable to the viewer.